Features of an effective thesaurus#

These were the features we were to use to evaluate each other’s thesauri:

• Vocabulary: terms are suitable.
• Reciprocality: each pair of terms is listed twice, so that both sides of the relationship are listed (BT/NT, RT/RT, and UF/USE).
• Logical coherence: effective use of different types of relationships (hierarchical vs related) and clear scope.
• Effect on search: the thesaurus would help the user in their goals.

Warrant behind terminology#

I spent a lot of time thinking about warrant for the hierarchical categories. For example, it seems to me that there is an essential difference between “Interpersonal emotions” like love or hate, and “Personal emotions” like happiness or sadness. You could maybe call these theory-based categories. But I chose to merge them into one category of “Emotions” to reflect the way that I think users might think about them; in other words, I chose to prioritize user warrant.

Another example of thinking about warrant is in the case of “Basic plots”. I used Christopher Booker’s (2004) Seven Basic Plots, as described by Wikipedia (“The Seven Basic Plots”, 2025). I felt that plots would be important to creators looking for storytelling inspiration. I decided to use these specific terms, which could be said to have scientific warrant, rather than something that would more closely resemble everyday language, because I thought that the users of this thesaurus would probably know something (or be interested in learning) about the theory/research behind plots. If I had made this for any person looking for a movie to watch, I probably would have prioritized user warrant and chosen terms that more closely resemble how the general public talks about plot.

Understanding reciprocality in an intuitive way#

Based on the thesauri that I peer reviewed, I think I grasped the idea of reciprocality more easily than others. Maybe this is because I have always enjoyed reading indexes and glossaries in non-fiction books. When reading a textbook about a new subject, I like to look at the index to see which concepts have been listed with the most granularity and the most page references; this helps give me a broad view of the field more effectively than reading the book itself.

I know idexes and glossaries are not the same as thesauri, but in a way, they share the principle of reciprocality in that they usually have preferred terms for synonyms. For example, an index/glossary might have the entry “food” with a definition (in a glossary) or page references (in an index), but then the entry “nourishment” may say “see ‘food’”. If ‘nourishment’ were only listed under “food”, then the user would not find it unless they thought to look under “food” (or they could use a ‘search’ function on a digital glossary/index). The relationship is often not truly reciprocal, in that they don’t often list “nourishment” under “food” as well. But the important thing is that someone who is looking for “nourishment” will find it in its alphabetical place and be directed to “food”.

It seemed to me that my peers often had an idea that some terms were more or less important or broad, and it was the more important/broad terms that got their own entries, whereas the narrower/minor terms were only ever listed under one or more broad terms. I’m guessing that full reciprocality can seem unnecessarily redundant and verbose if you don’t keep in mind that someone might be looking for each individual term in its alphabetical place.

Extra challenge: Adding color to see big-picture patterns#

I decided to add color to the main categories of terms so that I could try to see patterns in their groupings. The broadest terms are “Basic Plot”, “Relationships”, and “Emotions”. I also gave a different color to the “Love” category under “Emotions”, since that acted as its own broad term. Therefore, most of my thesaurus had only two levels, with the exception being “Love”.

One thing I noticed from the coloring was that there were associative relationships between the terms under “Relationships”, “Emotions”, and “Love”, but no connections between the terms under “Basic Plot” and other categories (i.e., the green terms are always isolated from the other-colored terms). Perhaps as a future improvement, I could do more research into the basic plot types and see if any types of relationships or emotions are consistently associated with plot types. For example, maybe the “Sadness” emotion is consistently associated with the “Tragedy” plot type, or “Happiness” with “Rags to Riches”. This might help users not only learn the conventional combinations of plot types and emotions, but also imagine unusual/creative combinations, like a “Rags to Riches” plot where the dominant emotion is “Anger”.

Change log between my original submission and this version#

• Fixed 1 mistake in alphabetization
• Fixed 1 missing reciprocal relationship
• Removed “offspring” as a preferred term for “children”, since the peer-reviewers thought it not likely that users would prefer “offspring”
• Merged “interpersonal emotions” and “personal emotions”, since the division seemed confusing to peer reviewers